OpenAI trial lawyer Sarah Eddy has been methodically dissecting Elon Musk’s claim of breach of charitable trust through the testimony of OpenAI President Greg Brockman.
Brockman is the source of the most damaging evidence to OpenAI by contemporaneous diary notes he took during the relevant period. For example:
But, during his testimony elicited by Eddy, Brockman has painted a much different picture of what happened.
Brockman’s testimony indicated several key points:
(1) Elon Musk wanted “unilateral control” of OpenAI and a majority equity stake in a for-profit company discussed by the leaders at OpenAI;
(2) Elon Musk proposed that Tesla absorb OpenAI and fund it; and
(3) In a tense meeting in 2017, when Brockman and Ilya Sutskever didn’t agree to Musk’s demand for unilateral control, Brockman said Musk’s mood changed. Brockman feared at the end of the meeting that Musk was circling the table to hit him. Musk didn’t, but he grabbed a painting made by Sutskever of a Tesla and stormed out.
(4) Musk wanted to pursue AGI at Tesla.
(5) Musk agreed with the need to have a for-profit company that could raise capital in the billions to compete with Google Deepmind (and beat Demis Hassabis, whom Musk did not like).
In my opinion, Brockman sounded like a calm, credible witness. His portrayal of Musk as someone who desired “unilateral control” over OpenAI–but was denied it–probably will resonate with some of the advisory jurors.
Judge Gonzalez Rogers will decide the case based on an advisory jury opinion.